This study was performed in a Neonatal Intensive Care Unit at Imam Khomeini Hospital of Ahvaz Jundishapur University of Medical Sciences in Ahvaz, Iran, during 2018–2019. Premature infants with gestational age of 28 to 34 weeks who had respiratory distress syndrome and their respiratory distress score based on the Silverman-Anderson table was 6 and 7 during the first 6 hours of birth were enrolled [13–16].
Exclusion criteria included presence of major anomalies, airway anomaly, severe cardiovascular instability, respiratory distress secondary to severe asphyxia (Apgar score ≤ 3 at 1 and 5 minute and umbilical cord blood pH < 7.1), parental discontent, gestational age less than 28 weeks, cyanotic heart disease, meconium aspiration syndrome, diaphragmatic hernia, invasive mechanical ventilation started from the beginning of hospitalization, pulmonary hemorrhage, lack of effective spontaneous breathing, metabolic disease during hospitalization and respiratory problems due to neuromuscular diseases and sepsis [11–16].
All parents were required to complete and write an informed consent form before the neonates were enrolled in the study, according to the Ethics Committee of Jundishapur University of Medical Sciences (IR.AJUMS.REC.1397.365). Also, the present study was registered in the Iranian Clinical Trial Documentation Office on 10.9.2018 (IRCT: 2018082 1040847NI).
In this double-blind clinical trial, neonates were randomly divided into two groups of NDUOPAP and NCPAP.NDUOPAP group was considered group A and NCPAP group as group B. Based on the https:// www. Sealedenvelope.com / simple–randomizer / V1 / lists, the list was prepared. Six blocks were initially considered, including AABB, ABAB, ABBA, BABA, BAAB, BBAA and each block was assigned a code between 1 and 6. The statistical consultant randomly selected a number from 1 to 6 to create a random sequence and as a result, the infants were randomized into the two groups of A and B. Sample size was calculated by formula and according to the sample size of Zhou et al's [11] article, where the failure rates of non-invasive NDUOPAP and NCPAP treatment were 4.44% and 22.5%, respectively, 67 patients were studied in each group. Due to the probability of at least 10% sample attrition, 7 individuals were added to each group, resulting in a sample size of 148 (74 subjects per group). After birth, the necessary resuscitation procedures were performed by a trained resuscitation team and a senior physician assistant for all infants who weighed below 1500 g according to the NICU protocol and infants were transited to NICU in presence of a specialized NICU nurse under T-piece respiratory support (Fisher & paykel Healthcare, New Zealand) [16].
In the NICU, infants who were eligible for inclusion were randomly assigned to one of NDUOPAP or NCPAP groups. In infants of the DUOPAP group Fabian device (Acutronic, Switzerland, Infant flow driver) were used, which was connected to the infant via standard nasal tubes and injectors through a nasal prong. For neonates in this group, baseline parameters including PDuo (8 cm H2O) and CPAP (5 cm H2O), FIO2 40%, inhalation time of 0.5 second, and respiratory rate between 30 to 40 breaths per minute were adjusted. Based on clinical examination, arterial blood gas (ABG) and SPO2, device parameters were changed. The highest acceptable CPAP and PDuo levels were less than 8 cm H2O and 15 cm H2O, respectively, and the maximum FIO2 acceptable to continue treatment was 60%. The goal of altering device setting was reaching SPO2 above 90% in the right hand, PaO2 above 50 mmHg, PaCO2 less than 50 mmHg, pH above 7.25 and lack of respiratory distress on physical examination [11, 13].
In the NCPAP group, infants were subjected to Fabian device (Acutronic, Switzerland, Infant flow driver). The device was connected to the infant by standard injectors and tubes through the nasal prong. In the NCPAP group the initial parameters of the device were CPAP (5 cm H2O) and FIO2 40% and based on clinical examination, ABG and SPO2 changes of device parameters were performed. The highest acceptable CPAP level was less than or equal to 8 cm H2O and the maximum FIO2 acceptable to continue treatment was 60%. The target was O2saturation above 90% in the right hand (PaO2 ≥ 50 cm H2 O, PaCO2 < 50 cm H2 O, and pH ≥ 7.25) and the absence of respiratory distress on physical examination [11, 13].
In both groups, based on existing therapeutic guides and under the direct supervision of the researcher, infants requiring FIO2 over 40% with CPAP > 5 cm H2O to maintain O2saturation in the right hand between 90 and 95%, 100 mg /kg surfactant (Survanta) were administered using the INSURE (Intubation, Surfactant and Extubation) method by a skilled practitioner who had been predetermined [17]. After INSURE, the infant received the same non-invasive ventilation used before INSURE.
A feeding tube was inserted to remove air from the baby's stomach. O2saturation was monitored and recorded by pulse oximeter and respiratory rate, heart rate was monitored continuously, and blood pressure every 2 hours. In infants requiring a FIO2 greater than 40% to maintain SPO2 within the acceptable range (90–95%), surfactant was re-administered after 6 hours after the last surfactant administration and as needed for a full course of treatment (maximum of 4 doses).
ABG was measured on admission (all subjects), in cases in need of intervention, one hour after the intervention as well as every 12 hours thereafter, and before and after surfactant administration and the results were recorded in a special form. Based on the results an appropriate intervention was carried out when necessary [11, 16, 18, 19]. Occurrence of treatment failure as well as duration of intervention, pneumothorax, BPD, PDA, apnea, occurrence of death, IVH, duration of oxygen therapy, length of hospital stay and mean airway pressure were recorded every 6 hours in each group. As decided, after improvement in patient's condition and O2saturation maintenance for 6 hours, we went on to reduce the device settings, such that if in DUOPAP FIO2 was less than 30% and CPAP and PDuo were less than or equal to 3 cm of water and 5 cm of water, respectively, and the infant was breathing continuously and ABG was normal for 24 hours, the infant was disconnected from the apparatus and placed under oxyhood inhaling a mixture of air and oxygen with FIO2 30–40% and a flow of 5 to 10 liters per minute depending on the size of the hood and patient's O2saturation [11].
In the CPAP group if the parameters were reduced to (CAPA ≤ 3 cm H2O and FIO2 ≤ 30%) and the infant was breathing normally and ABG was normal for 24 hour, the infant was separated from CPAP and subjected to oxyhood and oxygen/air mixture with FIO2 between 30–40% and flow ranging from 5 to 10 liters per minute depending on the size of the hood and patient's O2saturation [11].
All of the participants received antibiotics, caffeine as prophylaxis for apnea of prematurity and appropriate fluid and electrolyte solutions.
The primary outcome was the need for endotracheal intubation within the first 72 hours of treatment. Treatment failure criteria included at least one of the following: pH < 7.2, PaCO2 > 60 mmHg, PaO2 < 50 mmHg with FIO2 > 60%, CPAP > 8 cm H2O in NCPAP group and PDuo > 15 cm H2O, CPAP > 8 cm H2O, and FIO2 > 60% in NDUOPAP group or worsening of the clinical condition (increased respiratory distress due to severe retraction) or prolonged apnea (stopping breathing for more than 20 seconds) or recurring apnea more than 2 times in 24 hours with cyanosis and bradycardia (PR ≤ 100 / min) requiring ventilation with a bag and mask [11, 13, 20].
Secondary outcomes included duration of non-invasive ventilation, duration of oxygen therapy, duration of hospitalization, occurrence of IVH, pneumothorax, BPD, PDA, apnea, and death. All patients underwent echocardiogram within 48 hours of birth and afterward if needed. Brain ultrasonography for diagnosing IVH was performed on the third and seventh days. Pneumothorax was diagnosed on the basis of chest x ray and transillumination [12].
Statistical analysis
In quantitative variables mean and standard deviation were used to describe the data in addition to median and interquartile range. Frequency and percentage were used to describe the data. Normality of the data was analyzed using Kolmogorov-Smirnov test and Q-Q chart. Data were analyzed using chi-square, Fisher’s exact test, t-test and Mann-Whitney test. Significance level was set at P-value less than 0.05. All analyses were performed using SPSS version 22.